2D transition metal dichalcogenides for energy-efficient two-terminal optoelectronic synaptic devices
2D transition metal dichalcogenides for energy-efficient two-terminal optoelectronic synaptic devices
Two-dimensional layered transition metal dichalcogenides (2D TMDCs), such as tungsten disulfide, molybdenum disulfide, compounds based on rhenium, and their heterostructures, have been used to fabricate artificial synaptic devices that combine memory, computation, and sensing in a single system. By using a combination of optoelectronic/electronic signal processing systems, these devices have demonstrated multi-state memory, pattern-recognition capabilities, biological synaptic behavior, and visual information processing. Their advanced scalability and integration potential render them ideal candidates for emerging neuromorphic memories in edge AI and wearable devices. Although ultra-low power consumption in neuromorphic vision systems in the range of femtojoules has been achieved, optimizing the materials’ quality and controlling the defect formation are still required to enhance their functionality and improve the devices’ performance. Improving the scalability of heterostructures and integrating many single devices in arrays operating as part of a neuromorphic system are paramount to their commercialization.
DTI-2: Explore
Georgiadou, Dimitra
84977176-3678-4fb3-a3dd-2044a49c853b
Babu, Roshni
fe7dbe92-1f6b-44e7-94de-4321ff68284a
Georgiadou, Dimitra
84977176-3678-4fb3-a3dd-2044a49c853b
Babu, Roshni
fe7dbe92-1f6b-44e7-94de-4321ff68284a
Georgiadou, Dimitra and Babu, Roshni
(2025)
2D transition metal dichalcogenides for energy-efficient two-terminal optoelectronic synaptic devices.
Device, [100805].
(doi:10.1016/j.device.2025.100805).
Abstract
Two-dimensional layered transition metal dichalcogenides (2D TMDCs), such as tungsten disulfide, molybdenum disulfide, compounds based on rhenium, and their heterostructures, have been used to fabricate artificial synaptic devices that combine memory, computation, and sensing in a single system. By using a combination of optoelectronic/electronic signal processing systems, these devices have demonstrated multi-state memory, pattern-recognition capabilities, biological synaptic behavior, and visual information processing. Their advanced scalability and integration potential render them ideal candidates for emerging neuromorphic memories in edge AI and wearable devices. Although ultra-low power consumption in neuromorphic vision systems in the range of femtojoules has been achieved, optimizing the materials’ quality and controlling the defect formation are still required to enhance their functionality and improve the devices’ performance. Improving the scalability of heterostructures and integrating many single devices in arrays operating as part of a neuromorphic system are paramount to their commercialization.
Text
Device-D-25-00029_Perspective_FINAL with Figures
- Accepted Manuscript
Text
1-s2.0-S2666998625001188-main
- Proof
More information
Accepted/In Press date: 26 March 2025
e-pub ahead of print date: 23 May 2025
Keywords:
DTI-2: Explore
Identifiers
Local EPrints ID: 503986
URI: http://eprints.soton.ac.uk/id/eprint/503986
ISSN: 2666-9986
PURE UUID: 0e4132f5-4186-4a5e-9d29-a30d6e33c91e
Catalogue record
Date deposited: 21 Aug 2025 04:19
Last modified: 22 Aug 2025 02:28
Export record
Altmetrics
Contributors
Author:
Roshni Babu
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics